0
0

【YOLOv7_0.1】网络结构与源码解析

文章目录

前言

论文地址
YOLOv7源码

下面对v0.1版本的整体网络结构及各个组件，结合源码和

train

文件夹中的

yolov7.yaml

配置文件进行解析。

整体网络结构

分解的yolov7.yaml

# parametersnc:80# number of classesdepth_multiple:1.0# model depth multiplewidth_multiple:1.0# layer channel multiple# anchorsanchors:-[12,16,19,36,40,28]# P3/8-[36,75,76,55,72,146]# P4/16-[142,110,192,243,459,401]# P5/32# yolov7 backbonebackbone:# [from, number, module, args][[-1,1, Conv,[32,3,1]],# 0[-1,1, Conv,[64,3,2]],# 1-P1/2      [-1,1, Conv,[64,3,1]],[-1,1, Conv,[128,3,2]],# 3-P2/4# ELAN1[-1,1, Conv,[64,1,1]],[-2,1, Conv,[64,1,1]],[-1,1, Conv,[64,3,1]],[-1,1, Conv,[64,3,1]],[-1,1, Conv,[64,3,1]],[-1,1, Conv,[64,3,1]],[[-1,-3,-5,-6],1, Concat,[1]],[-1,1, Conv,[256,1,1]],# 11# MPConv[-1,1, MP,[]],[-1,1, Conv,[128,1,1]],[-3,1, Conv,[128,1,1]],[-1,1, Conv,[128,3,2]],[[-1,-3],1, Concat,[1]],# 16-P3/8# ELAN1[-1,1, Conv,[128,1,1]],[-2,1, Conv,[128,1,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[[-1,-3,-5,-6],1, Concat,[1]],[-1,1, Conv,[512,1,1]],# 24# MPConv[-1,1, MP,[]],[-1,1, Conv,[256,1,1]],[-3,1, Conv,[256,1,1]],[-1,1, Conv,[256,3,2]],[[-1,-3],1, Concat,[1]],# 29-P4/16# ELAN1[-1,1, Conv,[256,1,1]],[-2,1, Conv,[256,1,1]],[-1,1, Conv,[256,3,1]],[-1,1, Conv,[256,3,1]],[-1,1, Conv,[256,3,1]],[-1,1, Conv,[256,3,1]],[[-1,-3,-5,-6],1, Concat,[1]],[-1,1, Conv,[1024,1,1]],# 37# MPConv[-1,1, MP,[]],[-1,1, Conv,[512,1,1]],[-3,1, Conv,[512,1,1]],[-1,1, Conv,[512,3,2]],[[-1,-3],1, Concat,[1]],# 42-P5/32# ELAN1[-1,1, Conv,[256,1,1]],[-2,1, Conv,[256,1,1]],[-1,1, Conv,[256,3,1]],[-1,1, Conv,[256,3,1]],[-1,1, Conv,[256,3,1]],[-1,1, Conv,[256,3,1]],[[-1,-3,-5,-6],1, Concat,[1]],[-1,1, Conv,[1024,1,1]],# 50]# yolov7 headhead:[[-1,1, SPPCSPC,[512]],# 51[-1,1, Conv,[256,1,1]],[-1,1, nn.Upsample,[None,2,'nearest']],[37,1, Conv,[256,1,1]],# route backbone P4[[-1,-2],1, Concat,[1]],# ELAN2[-1,1, Conv,[256,1,1]],[-2,1, Conv,[256,1,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[[-1,-2,-3,-4,-5,-6],1, Concat,[1]],[-1,1, Conv,[256,1,1]],# 63[-1,1, Conv,[128,1,1]],[-1,1, nn.Upsample,[None,2,'nearest']],[24,1, Conv,[128,1,1]],# route backbone P3[[-1,-2],1, Concat,[1]],# ELAN2[-1,1, Conv,[128,1,1]],[-2,1, Conv,[128,1,1]],[-1,1, Conv,[64,3,1]],[-1,1, Conv,[64,3,1]],[-1,1, Conv,[64,3,1]],[-1,1, Conv,[64,3,1]],[[-1,-2,-3,-4,-5,-6],1, Concat,[1]],[-1,1, Conv,[128,1,1]],# 75# MPConv Channel × 2[-1,1, MP,[]],[-1,1, Conv,[128,1,1]],[-3,1, Conv,[128,1,1]],[-1,1, Conv,[128,3,2]],[[-1,-3,63],1, Concat,[1]],# ELAN2[-1,1, Conv,[256,1,1]],[-2,1, Conv,[256,1,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[[-1,-2,-3,-4,-5,-6],1, Concat,[1]],[-1,1, Conv,[256,1,1]],# 88# MPConv Channel × 2[-1,1, MP,[]],[-1,1, Conv,[256,1,1]],[-3,1, Conv,[256,1,1]],[-1,1, Conv,[256,3,2]],[[-1,-3,51],1, Concat,[1]],# ELAN2[-1,1, Conv,[512,1,1]],[-2,1, Conv,[512,1,1]],[-1,1, Conv,[256,3,1]],[-1,1, Conv,[256,3,1]],[-1,1, Conv,[256,3,1]],[-1,1, Conv,[256,3,1]],[[-1,-2,-3,-4,-5,-6],1, Concat,[1]],[-1,1, Conv,[512,1,1]],# 101[75,1, RepConv,[256,3,1]],[88,1, RepConv,[512,3,1]],[101,1, RepConv,[1024,3,1]],[[102,103,104],1, IDetect,[nc, anchors]],# Detect(P3, P4, P5)]

各组件结构

ELAN1 (backbone)

  • yolov7.yaml中对应部分：
# ELAN1[-1,1, Conv,[64,1,1]],[-2,1, Conv,[64,1,1]],[-1,1, Conv,[64,3,1]],[-1,1, Conv,[64,3,1]],[-1,1, Conv,[64,3,1]],[-1,1, Conv,[64,3,1]],[[-1,-3,-5,-6],1, Concat,[1]],[-1,1, Conv,[256,1,1]],# 11

ELAN2 (head)

  • yolov7.yaml中对应部分：
# ELAN2[-1,1, Conv,[256,1,1]],[-2,1, Conv,[256,1,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[-1,1, Conv,[128,3,1]],[[-1,-2,-3,-4,-5,-6],1, Concat,[1]],[-1,1, Conv,[256,1,1]],# 63

MPConv

  • 在backnone中的对应部分
  • 要注意相比于MP函数之前，通道数减少一半
[-1,1, Conv,[256,1,1]],# 11# MPConv[-1,1, MP,[]],[-1,1, Conv,[128,1,1]],[-3,1, Conv,[128,1,1]],[-1,1, Conv,[128,3,2]],[[-1,-3],1, Concat,[1]],# 16-P3/8
  • 在head中的对应部分
  • 要注意相比于MP函数之前，通道数不变
[-1,1, Conv,[128,1,1]],# 75# MPConv Channel × 2[-1,1, MP,[]],[-1,1, Conv,[128,1,1]],[-3,1, Conv,[128,1,1]],[-1,1, Conv,[128,3,2]],[[-1,-3,63],1, Concat,[1]],

SPPCSPC

类似于

yolov5

中的

SPPF

，不同的是，使用了5×5、9×9、13×13最大池化。

  • common.py中对应部分：
classSPPCSPC(nn.Module):# CSP https://github.com/WongKinYiu/CrossStagePartialNetworksdef__init__(self, c1, c2, n=1, shortcut=False, g=1, e=0.5, k=(5,9,13)):super(SPPCSPC, self).__init__()
        c_ =int(2* c2 * e)# hidden channels
        self.cv1 = Conv(c1, c_,1,1)
        self.cv2 = Conv(c1, c_,1,1)
        self.cv3 = Conv(c_, c_,3,1)
        self.cv4 = Conv(c_, c_,1,1)
        self.m = nn.ModuleList([nn.MaxPool2d(kernel_size=x, stride=1, padding=x //2)for x in k])
        self.cv5 = Conv(4* c_, c_,1,1)
        self.cv6 = Conv(c_, c_,3,1)
        self.cv7 = Conv(2* c_, c2,1,1)defforward(self, x):
        x1 = self.cv4(self.cv3(self.cv1(x)))
        y1 = self.cv6(self.cv5(torch.cat([x1]+[m(x1)for m in self.m],1)))
        y2 = self.cv2(x)return self.cv7(torch.cat((y1, y2), dim=1))

RepConv（重参数卷积）

原理理解层面

  • 训练时：一个33卷积、一个11卷积和一个BN层（当输入输出通道相同时）相加得到输出
  • 推理时：将以上三部分重参数化，合并为一个3*3的卷积输出

代码实现层面

  • 训练时：不执行Model类的fuse函数
  • 推理时：在attempt_load函数加载训练好的模型时，会执行Model类的fuse函数，进而调用fuse_repvgg_block函数，实现将三个卷积重参数化，合并为一个卷积输出
  • common.py中对应部分：
# Represented convolution https://arxiv.org/abs/2101.03697classRepConv(nn.Module):'''重参数卷积
    训练时：
        deploy = False
        rbr_dense(3*3卷积) + rbr_1x1(1*1卷积) + rbr_identity(c2 == c1时) 三者相加
        rbr_reparam = None
    推理时：
        deploy = True
        rbr_reparam = Conv2d
        rbr_dense = None
        rbr_1x1 = None
        rbr_identity = None
    '''def__init__(self, c1, c2, k=3, s=1, p=None, g=1, act=True, deploy=False):super(RepConv, self).__init__()

        self.deploy = deploy
        self.groups = g
        self.in_channels = c1
        self.out_channels = c2

        assert k ==3assert autopad(k, p)==1

        padding_11 = autopad(k, p)- k //2

        self.act = nn.SiLU()if act isTrueelse(act ifisinstance(act, nn.Module)else nn.Identity())# 推理阶段，仅有一个3×3的卷积来替换if deploy:
            self.rbr_reparam = nn.Conv2d(c1, c2, k, s, autopad(k, p), groups=g, bias=True)else:# 训练阶段，当输入和输出的通道数相同时，会在加一个BN层
            self.rbr_identity =(nn.BatchNorm2d(num_features=c1)if c2 == c1 and s ==1elseNone)# 3×3的卷积(padding=1)
            self.rbr_dense = nn.Sequential(
                nn.Conv2d(c1, c2, k, s, autopad(k, p), groups=g, bias=False),
                nn.BatchNorm2d(num_features=c2),)# 1×1的卷积
            self.rbr_1x1 = nn.Sequential(
                nn.Conv2d(c1, c2,1, s, padding_11, groups=g, bias=False),
                nn.BatchNorm2d(num_features=c2),)defforward(self, inputs):ifhasattr(self,"rbr_reparam"):return self.act(self.rbr_reparam(inputs))if self.rbr_identity isNone:
            id_out =0else:
            id_out = self.rbr_identity(inputs)return self.act(self.rbr_dense(inputs)+ self.rbr_1x1(inputs)+ id_out)# Conv2D + BN -> Conv2Ddeffuse_conv_bn(self, conv, bn):

        std =(bn.running_var + bn.eps).sqrt()
        bias = bn.bias - bn.running_mean * bn.weight / std

        t =(bn.weight / std).reshape(-1,1,1,1)
        weights = conv.weight * t

        bn = nn.Identity()
        conv = nn.Conv2d(in_channels=conv.in_channels,
                         out_channels=conv.out_channels,
                         kernel_size=conv.kernel_size,
                         stride=conv.stride,
                         padding=conv.padding,
                         dilation=conv.dilation,
                         groups=conv.groups,
                         bias=True,
                         padding_mode=conv.padding_mode)

        conv.weight = torch.nn.Parameter(weights)
        conv.bias = torch.nn.Parameter(bias)return conv

    # 在推理阶段才执行重参数操作deffuse_repvgg_block(self):if self.deploy:returnprint(f"RepConv.fuse_repvgg_block")

        self.rbr_dense = self.fuse_conv_bn(self.rbr_dense[0], self.rbr_dense[1])
        self.rbr_1x1 = self.fuse_conv_bn(self.rbr_1x1[0], self.rbr_1x1[1])
        rbr_1x1_bias = self.rbr_1x1.bias
        # self.rbr_1x1.weight [256, 128, 1, 1]# weight_1x1_expanded [256, 128, 3, 3]
        weight_1x1_expanded = torch.nn.functional.pad(self.rbr_1x1.weight,[1,1,1,1])# Fuse self.rbr_identityif(isinstance(self.rbr_identity, nn.BatchNorm2d)orisinstance(self.rbr_identity,
                                                                        nn.modules.batchnorm.SyncBatchNorm)):# print(f"fuse: rbr_identity == BatchNorm2d or SyncBatchNorm")
            identity_conv_1x1 = nn.Conv2d(
                in_channels=self.in_channels,
                out_channels=self.out_channels,
                kernel_size=1,
                stride=1,
                padding=0,
                groups=self.groups,
                bias=False)
            identity_conv_1x1.weight.data = identity_conv_1x1.weight.data.to(self.rbr_1x1.weight.data.device)
            identity_conv_1x1.weight.data = identity_conv_1x1.weight.data.squeeze().squeeze()# print(f" identity_conv_1x1.weight = {identity_conv_1x1.weight.shape}")
            identity_conv_1x1.weight.data.fill_(0.0)
            identity_conv_1x1.weight.data.fill_diagonal_(1.0)
            identity_conv_1x1.weight.data = identity_conv_1x1.weight.data.unsqueeze(2).unsqueeze(3)# print(f" identity_conv_1x1.weight = {identity_conv_1x1.weight.shape}")

            identity_conv_1x1 = self.fuse_conv_bn(identity_conv_1x1, self.rbr_identity)
            bias_identity_expanded = identity_conv_1x1.bias
            weight_identity_expanded = torch.nn.functional.pad(identity_conv_1x1.weight,[1,1,1,1])else:# print(f"fuse: rbr_identity != BatchNorm2d, rbr_identity = {self.rbr_identity}")
            bias_identity_expanded = torch.nn.Parameter(torch.zeros_like(rbr_1x1_bias))
            weight_identity_expanded = torch.nn.Parameter(torch.zeros_like(weight_1x1_expanded))# print(f"self.rbr_1x1.weight = {self.rbr_1x1.weight.shape}, ")# print(f"weight_1x1_expanded = {weight_1x1_expanded.shape}, ")# print(f"self.rbr_dense.weight = {self.rbr_dense.weight.shape}, ")

        self.rbr_dense.weight = torch.nn.Parameter(
            self.rbr_dense.weight + weight_1x1_expanded + weight_identity_expanded)
        self.rbr_dense.bias = torch.nn.Parameter(self.rbr_dense.bias + rbr_1x1_bias + bias_identity_expanded)

        self.rbr_reparam = self.rbr_dense
        # 前向推理时，使用重参数化后的 rbr_reparam 函数
        self.deploy =Trueif self.rbr_identity isnotNone:del self.rbr_identity
            self.rbr_identity =Noneif self.rbr_1x1 isnotNone:del self.rbr_1x1
            self.rbr_1x1 =Noneif self.rbr_dense isnotNone:del self.rbr_dense
            self.rbr_dense =None

ImpConv（隐性知识学习）


这一部分直接继承自YOLOR中的显隐性知识学习。一般情况下，将神经网络的浅层特征称为显性知识，深层特征称为隐性知识。而YOLOR的作者（同时也是YOLOv7的作者）则直接把神经网络最终观察到的知识称为显性知识，那些观察不到、与观察无关的知识称为隐性知识。

在

model/common.py

文件中，定义了两类隐性知识：

ImplicitA

和

ImplicitM

，分别对输入 相加 和 相乘：

# AddclassImplicitA(nn.Module):def__init__(self, channel, mean=0., std=.02):super(ImplicitA, self).__init__()
        self.channel = channel
        self.mean = mean
        self.std = std
        # 全0矩阵
        self.implicit = nn.Parameter(torch.zeros(1, channel,1,1))
        nn.init.normal_(self.implicit, mean=self.mean, std=self.std)defforward(self, x):# 全0矩阵 与 输入 相加return self.implicit + x

# MultiplyclassImplicitM(nn.Module):def__init__(self, channel, mean=0., std=.02):super(ImplicitM, self).__init__()
        self.channel = channel
        self.mean = mean
        self.std = std
        # 全1矩阵
        self.implicit = nn.Parameter(torch.ones(1, channel,1,1))
        nn.init.normal_(self.implicit, mean=self.mean, std=self.std)defforward(self, x):# 全1矩阵 与 输入相乘return self.implicit * x

训练时

在模型训练阶段，先对输入进行

ImplicitA

操作， 在进行1*1卷积，最后进行

ImplicitM

操作：

classIDetect(nn.Module):
    stride =None# strides computed during build
    export =False# onnx export
    end2end =False
    include_nms =Falsedef__init__(self, nc=80, anchors=(), ch=()):# detection layersuper(IDetect, self).__init__()
        self.nc = nc  # number of classes
        self.no = nc +5# number of outputs per anchor
        self.nl =len(anchors)# number of detection layers
        self.na =len(anchors[0])//2# number of anchors
        self.grid =[torch.zeros(1)]* self.nl  # init grid
        a = torch.tensor(anchors).float().view(self.nl,-1,2)
        self.register_buffer('anchors', a)# shape(nl,na,2)
        self.register_buffer('anchor_grid', a.clone().view(self.nl,1,-1,1,1,2))# shape(nl,1,na,1,1,2)
        self.m = nn.ModuleList(nn.Conv2d(x, self.no * self.na,1)for x in ch)# output conv# 初始化隐性知识
        self.ia = nn.ModuleList(ImplicitA(x)for x in ch)
        self.im = nn.ModuleList(ImplicitM(self.no * self.na)for _ in ch)defforward(self, x):# x = x.copy()  # for profiling
        z =[]# inference output
        self.training |= self.export
        for i inrange(self.nl):# 加入隐性知识
            x[i]= self.m[i](self.ia[i](x[i]))# conv
            x[i]= self.im[i](x[i])
            
            bs, _, ny, nx = x[i].shape  # x(bs,255,20,20) to x(bs,3,20,20,85)
            x[i]= x[i].view(bs, self.na, self.no, ny, nx).permute(0,1,3,4,2).contiguous()ifnot self.training:# inferenceif self.grid[i].shape[2:4]!= x[i].shape[2:4]:
                    self.grid[i]= self._make_grid(nx, ny).to(x[i].device)

                y = x[i].sigmoid()
                y[...,0:2]=(y[...,0:2]*2.-0.5+ self.grid[i])* self.stride[i]# xy
                y[...,2:4]=(y[...,2:4]*2)**2* self.anchor_grid[i]# wh
                z.append(y.view(bs,-1, self.no))return x if self.training else(torch.cat(z,1), x)

推理时

在模型推理阶段，将

ImplicitA-Conv-ImplicitM

融合为一个1*1的

Conv

操作：

# 将隐性知识与Detect层的1*1卷积进行融合deffuse(self):print("IDetect.fuse")# fuse ImplicitA and Convolutionfor i inrange(len(self.m)):
        c1, c2, _, _ = self.m[i].weight.shape
        c1_, c2_, _, _ = self.ia[i].implicit.shape
        self.m[i].bias += torch.matmul(self.m[i].weight.reshape(c1, c2),
                                       self.ia[i].implicit.reshape(c2_, c1_)).squeeze(1)# fuse ImplicitM and Convolutionfor i inrange(len(self.m)):
        c1, c2, _, _ = self.im[i].implicit.shape
        self.m[i].bias *= self.im[i].implicit.reshape(c2)
        self.m[i].weight *= self.im[i].implicit.transpose(0,1)

References

[1] 深入浅出 Yolo 系列之 Yolov7 基础网络结构详解
[2] 【yolov7系列】网络框架细节拆解
[3] yolov7-GradCAM

标签： 人工智能
本文转载自: https://blog.csdn.net/weixin_43799388/article/details/126164288
版权归原作者 嗜睡的篠龙 所有， 如有侵权，请联系我们删除。
登录后发布评论

“【YOLOv7_0.1】网络结构与源码解析”的评论:

还没有评论
关于作者
...
overfit同步小助手
文章同步
相关阅读

探索安全的Rust封装：ESP IDF SDK 的高效工具箱 —— esp-idf-svc

探秘GO Simple Tunnel：一款强大的Go语言安全隧道工具

Linux shell编程学习笔记78：cpio命令——文件和目录归档工具（上）

【DevOps】Maven 项目中常见的目录结构：集成测试代码的目录结构与单元测试的目录结构

Postman：Postman API测试：Postman中高级认证机制解析

Selenium驱动下载，浏览器最新版驱动下载

Mastercam V9.1软件安装教程、安装包下载

文章导航